Urban75 Home About Offline BrixtonBuzz Contact

autonomous cars - the future of motoring is driverless

How about a real-life Scalextric style system? That way you don't need to build more roads. It works for trams in many cities worldwide.

You mean like the Trams that are coming back in the UK maybe we could learn to do it properly again and have all the public transport modes joined up and working together again.....................like we used to have then less cars and less roads and possibly improve air quality
 
Could imagine top gears rage :mad::D
Some nissan sports car is already limited by gps go to a racing track in japan and you can go nuts
Otherwise the speed is limited.
 
Aside from Google there are other self driving car projects which use an adaptive approach and the cars process the data themselves. They start off with an initial set of data (map of the locality) and then use eg machine learning to refine this as they effectively explore their environment. Periodically the prior map is refreshed. Whole fleets of such cars could share data and learn from each other, continually fine tuning their map of the world, taking on new road layouts, roadworks, changes of priority, etc.

For example, the Nissan/Oxford RobotCar does this. I have noticed it on public roads but I don't know what mode it was in at the time. Possibly they were surveying for the prior map in advance of it being green flagged to self drive.
 
Did anyone see click last weekend? http://www.bbc.co.uk/iplayer/episode/b05y0fk1/click-cars
Cars that brake, complete with vibrating accelerator and will stop you tailgating. Cars that sense your mood or can read your journey ie home to work, work to gym, gym to home and it will remember your settings for tempreature, music Etc. They will read your mood, have vibrating seats.............the lists are endless! It's not driving anymore. I'm off to get myself a moggie or similar.
 
How self driving cars could be programmed to kill the occupants.
This 'dilemma that must be tackled' just seems like someone generating work for themselves. Why should a driverless vehicle take drastic and unpredictable evasive action instead of simply braking? How often does this scenario - an absolutely unavoidable and accidental collision with a large group of pedestrians - actually arise?
 
At the end of the programme about minis the other night ant went in a self driving Beemer in normal road conditions. It took evasive action by slowing, indicating and changing lanes all by itself when it saw a potential problem ahead. Scary stuff. When they are properly introduced, I suspect some people will see them as targets for fraudulent insurance claims.
 
I've yet to see a accident scenario where swerving is better than braking. Especially considering that computers are much better at braking than humans ever will be.
 
I've yet to see a accident scenario where swerving is better than braking. Especially considering that computers are much better at braking than humans ever will be.
A stationary vehicle on one lane of an otherwise empty motorway. You approach, see the vehicle in your lane, change lanes (swerve) to avoid a collision.
 
That AutoPilot is pretty amazing for motorway driving. I do a regularly 7 hour one day round-trip for work and 99% of that is on a single motorway. I could have a lovely sleep on the way there and back ;)

It would be awesome. As would it bringing you home from the pub.
 
A stationary vehicle on one lane of an otherwise empty motorway. You approach, see the vehicle in your lane, change lanes (swerve) to avoid a collision.

Equivocation. Changing lanes to avoid an obstacle seen in advance is different to the bollocks hypotheticals put forth, to support the notion that self-driving cars must make some sort of ethical choice rather than revert to a failsafe mode.
 
Equivocation. Changing lanes to avoid an obstacle seen in advance is different to the bollocks hypotheticals put forth, to support the notion that self-driving cars must make some sort of ethical choice rather than revert to a failsafe mode.
In normal driving drivers are often faced with dilemmas some small others more serious.

On the smaller side, say someone's pet cat runs out into the road in front of you while you are driving past a bus queue, a careful driver might check behind and if it was clear, brake to hopefully avoid killing the cat. But it might not be in order to swerve and potentially lose control so close to the bus queue just to save the cat.

Now imaging it was a child that ran out in front of the driver, while they are coming up to the bus queue and perhaps there was not enough time to stop, the wise driver might swerve away from the bus queue, toward the opposite side of the road, assuming there was no oncoming traffic. But what about if there was oncoming traffic, what then would be the most wise thing to do?

Hard enough for a human to decide the right course of action, and that is assuming there is enough of a moment to take a decision rather than act instinctively. Now for auto drive cars this must be programmed into the operating software, how hard is that going to be?
 
Could imagine top gears rage :mad::D
Some nissan sports car is already limited by gps go to a racing track in japan and you can go nuts
Otherwise the speed is limited.

Only the Japanese models and you can turn it off via the dashboard menu thing.
 
I don't want driverless cars, yet, though maybe when Im old and struggling with eyesight and other frailties. Would you need a current valid driving licence to be in "control" of a driverless car or could they be for use by everyone regardless of age and ability?
In the meantime, I guess you could input you destinantion into a car and using satelite technology, the car will take you to your destinantion via
the quickest, least congested route. Also, I guess there is no reason why delivery companies cannot use them for making a number
of drops at various points in one outing, which would be efficient.
But, I don't see them replacing/destroying high speed rail per this...
Driverless cars will be a boon for economies
"Driverless cars will destroy high-speed rail, usher in a new golden age for individual transportation and undermine cities by encouraging longer commutes."
 
I've just come back from a three day insurance conference, containing many of the top people in the industry. It was taken as read there that driverless cars will be with us soon and that they will effectively have replaced non-autonomous vehicles by 2050, with the difficult part being the intermission between the two extremes. This is no longer a matter for debate, it's accepted as fact.

The questions being discussed surrounded how liability will be handled, and the nature of the insurance (who, what, how) that would be needed to cover it.

(People weren't bothered about the need for motor insurance disappearing, by the way. As a whole, the U.K. motor insurance industry hasn't made a profit since 1994. Many would be quite happy to see the back of it.)
 
So far, innovations, safety features and the like have begun by being installed on luxury cars, seatbelts, ABS, ASR, air-bags, auto braking, anti-trap windows, after which they are slowly installed on more new vehicles. But even now, many years after its first installation many or even most cars don't have ABS.
 
Take airbags as an example, they were first patented in 1951 in the US, first installed in production cars in the mid-1970s, in 1981 in Germany in the Mercedes S-Class and now some 64 years after they were first patented they are still not 100% fitment on passenger cars.
 
The questions being discussed surrounded how liability will be handled, and the nature of the insurance (who, what, how) that would be needed to cover it.
Some of this is already cementing. Volvo have said that they intend to assume liability for whatever their vehicles do whilst in autonomous operation, and that if you're not willing to do that, you need to get out of the game.

(People weren't bothered about the need for motor insurance disappearing, by the way. As a whole, the U.K. motor insurance industry hasn't made a profit since 1994. Many would be quite happy to see the back of it.)
This is wheeled out a lot but is deeply conditional. The same line could be used to say that retail banking doesn't make much in the way of profit. It generates a load of cash flow that's used to invest.
 
Take airbags as an example, they were first patented in 1951 in the US, first installed in production cars in the mid-1970s, in 1981 in Germany in the Mercedes S-Class and now some 64 years after they were first patented they are still not 100% fitment on passenger cars.
Times have changed. It's comparatively easy to mandate equipment on mass production cars - look at the pace of EU emissions standards, for example. It's only harder to deal with legacy cars, but time will kill off most of them.

ABS is mandatory fitment on all new mass production cars in the EU, by the way, since 2004.
 
Last edited:
As for the wider topic, I only have glimpses of the state of the art, but I think the Tesla video shows that we're not there yet. How they get away with this 'beta' thing, I don't know, but it's clearly not ready for the real world. They seem to be heavily reliant on crowdsourced machine learning - where other cars have been - rather than on the fly analysis, which solves a lot of the 'problem' of driving a car, but will still be disastrous.

You can't do autonomous vehicles by halves - none of this interventionist human nonsense - and it's not clear that the infrastructure is going to change to accommodate it (e.g. at an extreme, segregated road networks) so I have my doubts about the technical feasibility in the next ten years or so. We'll see, I guess.
 
Will any of us actually be able to afford them. I suspect by the time it happens private motor transport will be a luxury for the rich only.
 
Back
Top Bottom